Prasanna H.
Master's Candidate
Experience
Apr 2024 - Nov 2024
8 monthsGermany
Master's Candidate
Robotics Research Lab, TU Kaiserslautern
- Benchmarked datasets, collected real-world off-road data (~9000 images) and generated simulation datasets using Unreal Engine.
- Developed Gen-AI image segmentation in Unreal Engine, reducing time from 1-2 days to 5-10 minutes.
- Built Generative AI data enhancement pipeline. Achieved an improvement in synthetic data by +49% mIoU.
- Tech: Python, PyTorch, C++, Git, LangChain, Gen-AI, Linux, W&B, OpenCV, Labelme.
Jan 2024 - Jul 2024
7 monthsUlm, Germany
Data Scientist Working Student
Bosch Rexroth
- Designed pilot backend ETL architecture with Azure (Cosmos DB, Blob Storage, Functions triggers) by combining and cleaning big data from different sources.
- Built frontend data analysis & visualization tools (PoC) with FastAPI, Python, Streamlit for semi-structured data.
- Leveraged Azure Functions to automate data pipeline, resulting in 25% reduction in time over manual data pipeline.
- Developed FastAPI endpoints for cloud data retrieval & storage and delivered production-ready code with Git and CI/CD.
- Developed Power BI dashboards to collaborate across teams to offer better product decisions and services.
- Tech: Python (Pandas, matplotlib), Azure (Cosmos DB, Blob, Functions), FastAPI, SQL/NoSQL, Agile, Kanban.
Jun 2023 - Nov 2023
6 monthsHomburg, Germany
Intern Data Scientist
Bosch Rexroth
- Cleaned, manipulated, and extracted features from production data for data analysis, visualization and supervised machine learning.
- Collected quality data and performed statistical analysis using Q-DAS qs-STAT to evaluate & enhance model performance.
- Predicted production part quality with machine learning, enabling real-time defect detection and eliminating 1–2-day delays from manual inspections.
- Automated Power BI dashboards, enabling real-time KPI monitoring and anomaly detection, reducing issue response time by 20%.
- Tech: Python (Pandas, NumPy, matplotlib), scikit-learn, MLflow, Power BI, Q-DAS.
Jul 2018 - Aug 2021
3 years 2 monthsIndia
Design Engineer
Actalent Engineering & Sciences Services
- Developed CAD parts & technical drawings with 98% first-time-right rate.
- Managed BOMs & PLM data in SAP and Siemens Teamcenter. Created technical documentation.
- Conducted risk analysis (FMEA) & automated part generation using Python NX journaling.
- Collaborated with stakeholders and suppliers to address critical design issues.
Chat With Jarvis
- Developed a custom LLM chatbot with RAG, supporting PDFs, DOCX and images; features were not available in popular LLMs at the time of project.
- Achieved over 95% monthly cost savings compared to Chat-GPT Plus.
- Tech: Python, LangChain, prompting, OpenAI API, Docker, Chroma.
Learning German with News
- Developed an LLM-based German learning app that delivers the latest news, translates it, and explains vocabulary & grammar, making language learning engaging and practical.
- Designed and implemented scalable AWS infrastructure using EC2 launch templates, IAM roles, ECR, Application Load Balancer, Auto Scaling Groups and CloudWatch to reliably deploy and manage the app.
- Tech: NLP, prompt engineering, Python, AWS (EC2, IAM, ECR, ECS, ALB, ASG).
Formula 1 Data Lakehouse Project
- Built ETL pipelines using PySpark and Spark SQL for raw data (CSV/JSON) ingestion & transformation, reducing manual processing time by 30%.
- Implemented lakehouse architecture with Delta Lake, supporting both incremental and full refresh ETL patterns.
- Tech: PySpark, Spark SQL, Delta Lake, Azure Data Lake Storage, Azure Data Factory, Databricks.
LLM Fine-tuning Language Translation
- Fine-tuned LLMs using LoRA and qLoRA techniques; created synthetic translation datasets for model training and evaluated model performance across multiple scenarios and metrics.
- Tech: NLP, prompt engineering, Huggingface Transformers, PEFT, TRL.
Wine Quality Prediction MLOps
- Worked on an end-to-end MLOps project using XGBoost, Optuna tuning and MLflow for experiment tuning & tracking.
- Containerized FastAPI backend and Streamlit UI with Docker; deployed on AWS ECS via CI/CD GitHub Actions, utilizing AWS S3 for storage and ALB for high availability.
- Tech: Python, XGBoost, Optuna, MLflow, FastAPI, Streamlit, Docker, AWS (ECS, S3, ALB), GitHub Actions.
Summary
A highly Analytical, curious & motivated Data Scientist with experience in ML, DL, and Generative AI. Proven experience in building data analysis tools, fine-tuning LLMs and LLM based application development. Experienced in API development, ETL, ELT, data visualization and cloud deployment (Azure, AWS). Strong mechanical engineering background with proven ability to apply data-driven solutions and statistical analysis to real-world challenges.
Languages
English
AdvancedHindi
AdvancedKannada
AdvancedGerman
IntermediateEducation
Oct 2021 - Nov 2024
TU Kaiserslautern
Master of Science, Commercial vehicle technology (Computer Science focused) · Commercial vehicle technology (Computer Science focused) · Kaiserslautern, Germany · 1.6
Aug 2014 - Jun 2018
NIE
Bachelor of Engineering, Mechanical Engineering · Mechanical Engineering · Mysuru, India · 1.51
Certifications & licenses
AWS Cloud Practitioner (CLF-C02)
Deep Learning Specialization
DeepLearning.AI
IBM Data Science Professional Certificate
IBM
Machine Learning Specialization
DeepLearning.AI
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.